The Psychology of Paranoid Delusions

Richard Bentall

Institute of Psychology, Health & SocietyLiverpool University

1: Defining delusions and the phenomenology of paranoia

Delusions are……

Bizarre or unusual beliefs. However, defining what makes a

particular belief delusional has taxed the minds of the greatest

psychopathologists…..

Karl Jaspers (1883-1960) argued that the

abnormal beliefs of psychiatric patients are

bizarre, resistant to counter-argument and held

with extraordinary conviction. However, true

delusions are also ununderstandable:

They cannot be understood in terms of

The patient’s personality

The patient’s experiences

And can therefore only be explained in terms of

aberrant biology.

Understandability

Berrios (1991): delusions are “empty speech acts, whose

informational content refers to neither world or self”.

DSM-IV (APA, 1994) defines a delusion as:

“A false personal belief based on incorrect inference about external reality that is firmly sustained despite what almost everyone else believes and despite what constitutes incontrovertible and obvious proof or evidence to the contrary. The belief is not one ordinarily accepted by other members of the person’s culture or subculture.”

A modern definition

Persecutory (paranoid) delusions

PANSS* > 2 N/255 %

Delusions (P1) 250 98

Suspicion (P6) 235 91.8

Delusions & Suspicion 230 90.2

Hallucinations (P3) 177 69.1

Thought disorder (P2) 144 56.5

Grandiosity (P5) 98 38.6

Agitation (P4) 179 70.2

Hostility (P7) 97 37.9

In the SoCRATES first episode sample (Moutoussis et al. 2007)

Positive and Negative Syndromes Scale (Kay et al., 1987), a widely used measure of

psychotic symptoms,

A paranoid continuum?

Many psychologists have argued that psychotic experiences exist on a

contiuum with normal functioning (e.g. Claridge, 1990) and have

developed psychometric instruments to assess this continuum (e.g.

Bentall, Claridge & Slade, 1988),

• Epidemiological studies show that large numbers of people

report delusional beliefs (12.0%, van Os et al., 2000, Holland) or

paranoia (12.6% paranoia, Poulton et al., 2000, New Zealand)

•Freeman et al. (2005) administered a paranoia questionnaire to

over 1000 people in a UK internet survey. They found evidence

for a continuum, although extreme beliefs about threats of harm

were only endorsed by a minority.

A paranoid continuum?

Proportion of population scoring on the Paranoia Checklist (from

Freeman et al. 2005).

A paranoid continuum?

Hierarchy of paranoia (from Freeman et al. 2005).

Two types of paranoia?

• Trower and Chadwick (1995) argue that there are two types of paranoia:

‘Poor me’ paranoia (persecution underserved, self-esteem preserved)

‘Bad me’ paranoia (persecution deserved, self-esteem low)

• However, there has been almost no research to examine the distinction.

The PADS (Melo et al., 2009)

Developed a Persecution and Deservedness scale (10, 12 and 20 item

versions) designed for both patient and clinical participants.

Administered to 312 undergraduates and 45 acutely psychotic patients.

1. There are times when I worry that others might be plotting

against me.

(ANS: 0 = certainly false; 4 = certainly true)

If you’ve answered 2 or above to the last question, please answer to

the following question:

1.1 Do you feel like you deserve others to plot against you?

(ANS: 0 = not at all; 4 = very much)

01

23

4

Deserv

edness

0 1 2 3 4

Paranoia

Students 95% CI

Patients 95% CI

The PADS (Melo et al., 2009)

Adequate reliability was found for both dimensions. In non-patients, a clear relationship was observed between paranoia and deservedness, but this relationship was absent in patients. In the patient sample, deservedness scores appeared to be suppressed.

Collip, Oorschot, Thewissen, Van Os, Bentall & Myin-

Germeys (2010)

Used a diary method (Experience Sampling) to examine variation of momentary paranoia and perceived social threat (“In this company I feel threatened”, “In this company I feel accepted” [reversed]) with social context:

whether alone, in unfamiliar company, in familiar company

subjective stress since last bleep

At low and moderate levels of trait paranoia (Fenigstein scale),paranoid thinking and perceived social threat were predicted by both context and subjective stress.

At high levels of trait paranoia, paranoid thinking and perceived social threat were NOT predicted by either context and subjective stress.

Fluctuations in deservedness (Melo et al., 2006)

43 paranoid patients compared with 22 healthy controls.

Initial intention was to repeat assessments of paranoid patients

after 1 month – proved difficult.

All patients completed a deservedness analogue scale on each

assessment “0 = I don’t deserve to be persecuted”; “12 = I deserve

to be persecuted”.

Fluctuations in deservedness (Udachina et al. in

press)

02

46

8D

eserv

edness

poor-me bad-me

6 7 102337414548778586437589 202529324751396974173056408466

ESM study with 14 PM and 15 BM patients. Deservedness

measured at each beep.

Time

Des

erved

nes

sTime course of deservedness?

Bad me

Poor me

Bad me

Poor me

Prodrome First episode

2: Evolutionary and developmental origins of paranoia

In the conditioned avoidance paradigm, the animal is placed in a shuttle box, in

which it can receive a warning signal and an electric shock.

All organisms need a system for anticipating and avoiding

threat

Aversive SAversive S EscapeEscapeWarning SWarning S

Avoidance RAvoidance RNo aversive SNo aversive SWarning SWarning S

Note that learning

continues (decreased

response latencies) long

after 100% avoidance is

achieved.

Does repeated exposure to threat lead to

psychosis?: A meta-analysis

Initial database search found 27,572 hits- after excluding studies based of inspection of the papers' titles and abstracts, the 763 remaining papers were examined for inclusion.

The analysis refers to studies focusing on EARLY adversity (exposure to trauma, bullying, parental death etc before the age of 18) and psychosis (both diagnostic and dimensional outcomes) with the following designs:

• epidemiological cross-sectional studies• prospective studies (and quasi prospective studies)• patient control studies

Study name Odds ratio and 95% CI

Bebbington et al., 2011 Epidemiological cross-sectional

Harley et al., 2010 Epidemiological cross-sectional

McAloney et al., 2009 Epidemiological cross-sectional

Nishida et al., 2008 Epidemiological cross-sectional

Shevlin et al., 2008 Epidemiological cross-sectional

Whitfield et al., 2005 Epidemiological cross-sectional

Epidemiological cross-sectional

Evans, 2011 Patient-control

Fisher et al., 2010 Patient-control

Habets et al., 2011 Patient-control

Husted et al., 2010 Patient-control

Rubino et al., 2009 Patient-control

Sommer et al., 2010 Patient-control

Stompe et al., 2006 Patient-control

Varese et al., 2011 Patient-control

Weber et al., 2008 Patient-control

Patient-control

Arseneault et al., 2010 Prospective (and quasi-prospective)

Cutjar et al., 2010 (M) Prospective (and quasi-prospective)

Cutjar et al., 2010 (F) Prospective (and quasi-prospective)

De Loore et al., 2007 Prospective (and quasi-prospective)

Schreier et al., 2009 Prospective (and quasi-prospective)

Spauwen et al., 2006 Prospective (and quasi-prospective)

Prospective (and quasi-prospective)

0.01 0.1 1 10 100

Increased likelihood

Association between trauma and psychosis

The findings suggest a significant association between trauma and psychosis across all different reserach designs (patient-control studies:

• patient-control studies: OR = 3.3

• epidemiological cross-sectional: OR = 2.5

• prospective: OR = 2.6

Does repeated exposure to threat lead to

psychosis?: A meta-analysis

Is there a specific association between victimization

and paranoia?

Mirowski and Ross (1981) reported data on paranoid beliefs from a

community survey of residents of El Paso and Juarez. Paranoia was

associated with an external locus of control and experiences of

victimization and powerlessness.

The high risk of psychosis in immigrant groups (Harrison et al.,

1988), especially those living in relative isolation from other

immigrants (Boydell et al. 2001) might be explained in this way.

Janssen et al. (2003), in an epidemiological study of 7000+ Dutch

citizens, found that experiences of discrimination predicted the later

development of paranoid symptoms.

Attachment: A developmental vulnerability?

Dozier at al. (1991, 1995) found that schizophrenia patients,

especially with paranoia, most likely to have dismissing-avoidant

attachment style.

Community surveys of 8000 adults (Mickleson et al., 1997) and

1500 adolescents (Cooper at al., 1998) also show psychosis,

especially paranoia, associated with insecure attachment.

Early separation from parents (Morgan et al. 2006) and being

unwanted at birth (Myhrman et al. 1996) increase the risk of

psychosis in later life.

Pickering, Simpson & Bentall (2008)

503 students completed online questionnaires:

• The PADS (Melo et al. in press)

• The Launay–Slade (1981) Hallucination Scale

• Bartholomew and Horowitz’s (1991) Relationship Questionnaire

• Levenson’s multidimensional locus of control scale

(Externality, Chance, Powerful Others)

• Positive and negative self-esteem (Nugent & Thomas, 1993)

• Anticipation of threatening events (Bentall et al. in press)

Insecure attachment predicted paranoia when hallucinations were controlled for:

R2 = .53, for model including attachment anxiety, negative self-esteem,

anticipation of future threat, the recall of threat and powerful others

Insecure attachment did not predict hallucinations when paranoia was controlled for.

Specificity of adversities for paranoia

Data from the 2007 Adult Psychiatric Morbidity Survey (N =

7000+), which has measures of psychotic symptoms, and different

kinds of childhood adversity.

3: Psychological mechanisms

Psychological processes that have been implicated

in paranoia

Jumping to conclusions (e.g. Garety et al. 2001):

Patients with delusions tend to ‘jump to conclusions’ (make a decision about

uncertain events) on the basis of little information

• Typically measured by ‘the beads task’

• Well replicated finding

• Seems to be associated with delusions rather than specifically paranoia

Jumping to conclusions bias

Huq, Garety & Hemsley (1988)

Participants shown 2 jars with beads of two colours,

in ratios of 80:20 and 20:80. A sequence of beads is

shown, apparently from one of the jars. Participants

had to guess which jar.

Participants with delusions tended to ‘jump to

conclusions’, guessing after fewer draws. Although

well-replicated (Young & Bentall, 1997)

Psychological processes that have been implicated

in paranoia

Jumping to conclusions (e.g. Garety et al. 2001):

Patients with delusions tend to ‘jump to conclusions’ (make a decision about

uncertain events) on the basis of little information

• Typically measured by ‘the beads task’

• Well replicated finding

• Seems to be associated with delusions rather than specifically paranoia

Theory of mind (e.g. Corcoran & Frith, 1996):

It has been argued that paranoid patients have difficulty in understanding other

people’s thoughts and feelings (they have a poor ‘theory of mind’)

• Assessed by false belief stories, hinting tasks or even appreciation of jokes

• Psychotic patients perform poorly on ToM tasks, but specificity to

paranoia is ot proven

Attributions (e.g. Kaney & Bentall, 1989):

People make explanations (‘attributions’) for noteworthy events many times in a single day. ‘Attributional style’ refers to individual differences in the way that we construct explanations. Eg. from the Attributional Style Questionnaire (ASQ):

You fail an exam. Write down one possible cause__________________

InternalInternal

Totally due to me

StableStable

I can’t change it

Unstable

I can change it

External

Due to others or

circumstances

GlobalGlobal

It will affect all areas of my life

Specific

It will only affect

examinations

Attributional (Explanatory) Style

Positive Negative

ParanoidDepressedNormal

Fundamental observation:Paranoid patients makeabnormal attributions

Internality Scores*

10

20

30

40

Eg. Kaney & Bentall (1989)using the ASQ, found thatparanoid patients madeexcessively stable and globalattributions for negativeevents. More importantly,they showed an extremeself-serving bias.

Attributional (Explanatory) Style

* High scores mean

self-blaming, low

scores mean the cause

is something to do

with other people or

circumstances.

3 loci of causal attribution?

Research suggests that the bipolar internality scale of Peterson et al's (1982) ASQ:

is in need of revision. Kinderman & Bentall (1996) have suggested a three-way categorization of the internality dimension :

Internal

Totally due to me

External

Personal

External

Situational

External

Totally due to other people or

circumstances

Internal Totally due to me

Totally due to another person

or other people

Totally due to the situation

(circumstances or chance)

Kinderman & Bentall (1997) examined the tendency to makefewer internalattributions for negative events, and the tendency to make personal as opposedto situational external attributions in paranoid, depressedand non-patientparticipants.

2

3

4

5

6

7

8

9

10

Paranoid

Depressed

Non-patient

Attributional loci

Internal Personal Situational

+ - + - + -

3 loci of causal attribution?

The original attributional model

Bentall, Kinderman & Kaney (1994) proposed that an externalizingattributional style minimizes accessibility of negative self-schemas at the expense of generating paranoid beliefs.

External (other-blaming) attribution

Reduced negative thoughts about self

Threat of activation of negative beliefs about self

Increased belief that others have malevolent intentions towards self.

Experimental Group

Depression (Younger)

Depression w

P.D

elus

Control

Rem

itted Paranoid

Paranoid

Mean

110

100

90

80

70

60

50

40

30

Neg self-esteem

Pos self-esteem

Wellcome Paranoia Study: Schizophrenic paranoid (N=38), remitted schizophrenic paranoid (N=27), depressed paranoid (N=18), depressed non-psychotic (N=27) and control participants (N=33) (Bentall et al., in press.)

Correlations between

negative self-esteem &

paranoia (Fenigstein Scale)

Spearman r

SZ-P .32

SZ-R .41*

DEP-P .42

DEP-NP .53*

Control .39*

Problem #1: The relationship between self-esteem and paranoia (Bentall et al., 2008)

• Patients with positive psychotic symptoms (n=79), individuals with an at-risk mental state for paranoid psychosis (n=38), and control subjects (n=38) assessed using experience sampling method (ESM).

• 6 day diary, 10 bleeps/day:– 4 items measuring momentary self-esteem

– Other items measuring context, significant

experiences and attributions

Thewissen, Bentall, Lecomte, van Os & Myin-Germeys

(2008)

Thewissen, Bentall, Lecomte, van Os & Myin-Germeys

(2008)

SE instability Confounders n β (SE) p

Momentary level 1 -

sex, depression

sex, depression, SE level

155

155

155

.11 (.03)

.10 (.03)

.09 (.03)

p<.001

p<.01

p<.01

Day level 1 -

sex, depression

sex, depression, SE level

155

155

155

.21 (.06)

.20 (.07)

.17 (.07)

p<.01

p<.01

p=.01

Subject level 2 -

sex, depression

sex, depression, SE level

155

155

155

.13 (.03)

.13 (.04)

.11 (.03)

p<.001

p<.001

p<.01

Paranoia was associated with average low self-esteem, an effect that survived correction for depression but not SE instability. More importantly paranoia also independently related to SE stability:

1 Multilevel linear random regression model, β can be interpreted identically to the regression outcome in a unilevel linear regression model.

2 Unilevel linear regression model * p<0.05; ** p<.01; *** p<.001

Fluctuations and PM vs BM (Udachina et al. in

prep)

ESM study with 14 PM and 15 BM patients. 15 remitted patients

and 23 controls also assessed. Self-esteem and deservedness

measured at each beep.

• Deservedness was predicted by concurrent self-esteem.

• Onset of paranoia was predicted by drop in self-esteem from the

previous bleep.

• In BM patients, increases in paranoia were followed by decreases

in self-esteem at the next beep.

• In PM patients, remitted patients and controls, increases in

paranoia were followed by increases in self-esteem at the next beep.

Grey, Evans, Valiente & Bentall (in prep)

Two studies have reported that paranoid patients sometimes show low implicit self-esteem but relatively preserved explicit self-esteem (Moritz & Woodward, 2005; McKay et al. 2005). We measured implicit (Implicit Attitudes Test) vs explicit SE (Nugent & Thomas’s scale) in poor-me patients, bad-me patients and controls.

PM = BM < C All groups sig different

Associations between attributions and clinical paranoia have a ‘now-you-see-it, now-you-don’t’ aspect:

Replications:

• Candido & Romney (1990) (Canada)

• Fear et al. (1996) (Wales)

• Lassar & Debbelt (1998) (Germany)

• Lee & Wong (1998) (South Korea)

Partial replications:• Kristev et al. (1999) (Australia; partial replication)• Martin & Penn (2002)• McKay et al. (2005)

Problem #2: Is the association between attributions and paranoia replicable?

Complete failures to replicate:• Humphries and Barrowclough (2006)

Attributional abnormalities present in acute paranoid but not ‘normal’ paranoids• Jannsen et al (2006)• McKay et al. (2005)• Martin & Penn (2001 – non-patients) vs Martin & Penn (2002 – patients)

Attributions and deservedness (Melo et al., 2006)

p < .001

p < .01

* Low attribution scores indicate external attributions for negative events

Take home message from Part 3

• Negative self-esteem seems to be a very important factor in

paranoia.

• Excessively external attributions for negative events are only

found in acutely ill poor-me patients when they are ‘poor me’

• Self-esteem is also highly unstable in paranoia (possibly

relating the dynamic transitions between poor-me and bad-me

beliefs)

• Poor me paranoia is also associated with discrepancies

between implicit and explicit self-esteem.

4: Are all of the theories correct?

Must everyone get prizes?

In our recent Wellcome Trust funded study we combined data from the

following groups (Bentall et al. 2009):

• Schizophrenia patients with paranoid delusions

• Schizophrenia patients with paranoid delusions in remission

• Depressed patients with paranoid delusions

• Depressed patients without paranoid delusions

• Patient with late onset (aged => 65) schizophrenia-like psychosis with

paranoid delusions

• Elderly (aged => 65) depressed patients without paranoid delusions

• Healthy controls

Must everyone get prizes?

And modelled the relationships between measures of:

• Paranoid beliefs

• Threat anticipation

•Attributional style (excluding internality)

• Self-esteem (positive and negative)

• Depression and anxiety

• Theory of mind (2 measures)

• Jumping to conclusions (2 measures)

• Cognitive (executive) function (short WAIS and digit span backwards)

Could all of these theories be true?

Structural equation modelling revealed the following relationships:

THR1 THR2 THR3 THR5THR4 THR6 THR7

DFO1 DFO2 DFO3

GLOB

STAB

ST1

GL1

ST2

ST3

ST4

ST5

ST6

GL2

GL3

GL4

GL5

GL6

EMDYS

ANX DEP SEp SEn

INTFUN

IQv IQm DIGIT

ANTICPARB

ToMD

PDI2

PDI3

PDI4

PDI1

JTC

BIJ1

BIJ2

BIJ3

SOC1

SOC2

SOC3

.65 .81 .66.62 .68.80

.38

.59

.48

.46

.55

.70

.42

.55

.53

.42

.38

.58.97

.95

.92

.76

.87 .71 -.62 .92

.72

.78

.60

.46

.82

.60 .87 .93 .70 .64 .65 .55

.84

.90

.78

.92

.92

.88

PARCOG_P DEP_S

.92 .85

.72

-.39 .68

-.07

.25

.02

.17

.08

.22

.17

DSO2

DSO1

DSO3

.86

.32

.55

Could all of these theories be true?

Structural equation modelling revealed the following relationships:

Globality

Stability

Negative

emotionExec

function

Threat

AnPDI

ToM

JTC

.72

.78

.60

.46

.82

ParanoiaCognitive

Perfom.

pessimism

.92 .85

.72

-.39 .68

-.07

Take home message from Part 4

• There is evidence to support the role of multiple psychological

processes in paranoia

• These can be broadly grouped into two classes: emotional

(self-esteem and attributions) and cognitive (executive

function?)

• Emotional factors seem to be more important

• BUT the idea of a paranoid defence seems to still have some

mileage with respect to poor-me delusions in acutely ill patients

5: Some biological speculations

The animal is placed in a shuttle box, in which it can receive a warning signal and

an electric shock.

The conditioned avoidance paradigm

Aversive SAversive S EscapeEscapeWarning SWarning S

Avoidance RAvoidance RNo aversive SNo aversive SWarning SWarning S

Note that learning

continues (decreased

response latencies) long

after 100% avoidance is

achieved.

Could attributional responses seen in poor-me paranoia be

construed as covert avoidance responses?

Aversive

emotional state

Aversive

emotional stateNegative

thought

Negative

thought

External

attribution

External

attributionNo aversive

emotional state

No aversive

emotional state

Negative

thought

Negative

thought

How does CAR relate to paranoia?

Dopamine-blocking drugs abolish the conditioned avoidance

response (CAR) in animals (Beninger et al., 1980; Smith et al. 2005),

but not escape responding – suggests a role for dopamine in threat

perception. Hence, the CAR has long been used as initial screen for

antipsychotic drug action.

The CAR and dopamine

Drugs which block d-2 receptors in the striatum have a powerful

therapeutic effect on patients who experience persecutory delusions.

Animal studies show that repeated exposure to social defeat in animals

leads to sensitization of the dopamine system (Selten, 2005).

Paranoia as the end point of a developmental

pathway

Threat

anticipation

Threat

anticipation ParanoiaParanoia

Victimisation/

powerlessness

Victimisation/

powerlessness

Insecure

attachment

Insecure

attachment

Abnormal

cognitive style

Abnormal

cognitive style

Psychological description:

Paranoia as the end point of a developmental

pathway

Abnormal

dopamine

Abnormal

dopamine ParanoiaParanoia

Victimisation/

powerlessness

Victimisation/

powerlessness

Insecure

attachment

Insecure

attachment

Abnormal

cognitive style

Abnormal

cognitive style

Biological description:

Conclusions

• There seems to be a discontinuity between moderate and

severe paranoia

Moderate Paranoid Severe Paranoia

Type of delusion Bad-me Poor-me, but switching to

bad-me

Context dependency Dependent Independent

Self-esteem

Implicit Low Low

Explicit Low Higher

Stability Less unstable Highly unstable

Attributions Normal External for negative

events

Biological mediator ? Hyper-dopaminergia?

Antipsychotic responsive No Yes

Conclusions

• It is possible that the biological mediator of some of these processes

is the striatal dopamine system

• Abnormal cognitive functioning, leading to paranoid ideas, many

be the consequence of particular types of adverse life experiences

(insecure attachment and victimization).

• Paranoia can be explained in terms of the interaction between a

relatively small number of cognitive and emotional processes

• Genetic speculation (possibly naive): The A1 allele of the DRD2 gene

is associated with lower striatal D2 receptor density and poor

avoidance learning (Klein et al. 2007) - it should therefore be

protective against paranoia!

Clinical implications and further directions

• It seems likely (but not certain at this stage) that transition to

psychosis from the prodromal state is associated with a switch

from bad-me to poor-me. Consistent with this, we have found

that low self-discrepancies (good self-esteem) is positively

associated with psychosis in a prodromal sample (Morrison et

al. 2006).

• We can hypothesize that the shift from bad-me to poor-me on

transition to psychosis (if it occurs) is a result of attempts to

avoid negative thoughts (consistent with the assumptions

behind Bach & Hayes’ (2002) Acceptance and Commitment

Therapy version of CBT for psychosis)

• This conjecture needs to be tested in future studies:

- Longitudinal investigations of deservedness beliefs and

avoidance behaviours in high risk samples

- Experimental, genetic and neuroimaging studies of

avoidance learning in clinical samples

- Tests of new psychological interventions designed to

prevent abnormal avoidance behaviour.

Implications and further directions

That’s all folks!